Rosemary R C McEachan 1, Tiffany C Yang 2, Hannah Roberts 3, Kate E Pickett 4, Diana Arseneau-Powell 4, Christopher J Gidlow 5, John Wright 3, Mark Nieuwenhuijsen 6
Background: It is unknown whether the quantity or quality of green space is more important for mental wellbeing. We aimed to explore associations between availability of, satisfaction with, and use of green space and mental wellbeing among children aged 4 years in a multi-ethnic sample.
Methods: We did a 4-year follow-up assessment of participants in the Born in Bradford longitudinal cohort study, which recruited children and mothers at the city’s main maternity unit from 2007 to 2011. The primary outcome was parent-reported mental wellbeing for children aged 4 years, assessed with the standardised Strengths and Difficulties Questionnaire. Total, internalising, and externalising behavioural difficulties and prosocial behaviour scales were computed (with higher scores indicating greater difficulties or more prosocial behaviour). Residential green space around participants’ home addresses and distance to major green spaces were computed with the normalised difference vegetation index (NDVI). A subsample of participants completed additional questionnaires on measures of satisfaction with, and use of, local green spaces. Multiple regressions examined associations between green space and children’s mental wellbeing and explored moderation by ethnicity (white British vs south Asian) and socioeconomic status.
Findings: Between Oct 1, 2012, and June 30, 2015, 2594 mothers attended a follow-up appointment during which they completed a detailed questionnaire assessing the health of their child. 1519 (58%) participants were of south Asian origin, 740 (29%) of white British origin, and 333 (13%) of another ethnicity. Data on ethnicity were missing for two participants. 832 (32%) of 2594 participants completed additional questionnaires. Ethnicity moderated associations between residential green space and mental wellbeing (p<0·05 for total and internalising difficulties). After adjusting for all relevant variables, more green space was associated with fewer internalising behavioural difficulties (mean NDVI 100 m: β -2·35 [95% CI -4·20 to -0·50]; 300 m: -3·15 [-5·18 to -1·13]; 500 m: -2·85 [-4·91 to -0·80]) and with fewer total behavioural difficulties (100 m: -4·27 [-7·65 to -0·90]; 300 m: -5·22 [-8·91 to -1·54]; 500 m: -4·82 [-8·57 to -1·07]) only for south Asian children across all three buffer zones. In the subsample of participants, the effect of NDVI on wellbeing was rendered non-significant after controlling for satisfaction with, and use of, green space. Among south Asian children, satisfaction with green space was significantly associated with fewer total behavioural difficulties across all three buffer zones (β -0·59 [95% CI -1·11 to -0·07]), fewer internalising behavioural difficulties within 100 m (-0·28 [95% CI -0·56 to -0·003]) and 300 m buffer zones (-0·28 [-0·56 to -0·002]), and greater prosocial behaviour across all three buffer zones (0·20 [0·02 to 0·38]); no such associations were observed among white British children.
Interpretation: Positive effects of green space on wellbeing differ by ethnicity. Satisfaction with the quality of green space appears to be a more important predictor of wellbeing than does quantity of green space. Public health professionals and urban planners need to focus on both quality and quantity of urban green spaces to promote health, particularly among ethnic minority groups.
Funding: European Community's Seventh Framework Programme.
Open Access
